Umbraco 7 questions (many-to-many / lookup tables and more).
Hello people,
I am looking into what kind of things I can
build with Umbraco compared to using straight ASP.NET. I have tried to
understand the available documentation, but the lacking of quality and
mismanagement of the written content only gives small hints about what
is possible and what is not.
1) Many to many relationships (in a graph where many items can refer to the same). I want to create a repository system where everything can refer to everything bi-directional. E.g. like in a movie database: an actor can be the star in multiple movies, a movie can belong to multiple categories (action/drama). In a software system, you could have: a software package deployed on a host. In both cases, searching for a genre should suggest the movies and searching for a moviestar should show the movies that he or she is in. I have seen that for 1 to many relationships I could use the Relations API. However I did not find a clear answer to if it can be used to design Many-to-many solutions... or if this is recommended in Umbraco or possible. Umbraco has a tree-structure that does not look suitable if all nodes are referring to each other as in the graph data structure.
2) Look up tables - adding new values in one place for universal values. What is a common way to create a list of values that can be used to "group" things? I know I can use a Document Type as base, and create other Document Types under that so they get the value from its base. E.g. A list of pre-defined countries or departments.
3) Add tables in the Umbraco database and maybe integerate it in Umbraco. It is OK to use the database that Umbraco uses, and add my own tables which will be used in the solution? Will it have a bad impact when upgrading Umbraco?
4) Case with sharing data (need to reference the same document unless the IsNotShared is not on). E.g. in an Issue / Solution system. Let's say I have created a
single software product which is used by four teams. When a group finds
an issue and solution, they can add it to the software products' Issue /
Solution System. Now, if one of the groups make changes so it is a now a
more customized version, how it is possible for that group to add Issue
/ Solution items to the same software product, so THEY can see it, but
not the other groups who share the other common features? Each group
would have access to the same I/S system, but some of the other I/S
tickets would only be viewable for the group who creates it (unless they
want to show it).
It's hard for me to imagine a tree-structure in Umbraco that can do this... and new groups could be created on the fly...
5) Can the content node contain enough items for maybe a full-blown ticket system? I'm not thinking about the performance of the backoffice GUI when it needs to show nodes, more if the display of data to the user on the frontend will be affected?
6) Visual Studio and the media library - avoid copying the media folder. I can see that Umbraco saves images and flat files directly in the Umbraco folder. If more and more people add flat files and images to that folder it will take a long time to import the folder into Visual Studio. Are there any best practices for moving the media folder to another location and have Umbraco reference it? It would surely save some time when importing into VS.
This was a long post, I hope a friendly soul out there will have the time to give me an answer :)
1. There is a relations package and using the multi tree node picker you can relate things and do it bi directional its obviously not going to be as good as a graph db so if you really have loads of relations use a graph db, you have the data in umbraco so its easy for editors to edit then on publish using event handlers you can push that data to your graph db of choice im really a big fan of neo4j I have done any umbraco with it but Pete Duncanson who is quite popular in the umbraco world has done some prototyping.
2.You can do it with your own doc types however you can also do it csv files and use nupickers package, this will read the csv and give you a drop down we used it on a recent project to tag items with data like shoe size, hair colour etc.
3.You can add your own tables, upgrade will not trash those custom tables, also with upgrades db upgrades are done with migration scripts via peta poco.
4.This looks a bit complex so I do not fully understand it but if they create the issue you can add it to the product and using events set the permissions so this group can see and others cannot.
5. You can show the data on front end in paged manner? You can also on front end retrieve the data using Examine / Lucene searches so its lightening quick
6. There are amazon s3 providers that you can use and swap out the out of the box media provider, we are doing this on a current project as we have over 50gb of media so it all goes up into the cloud as blob storage. This is done by modifying the FileSystemProdiers.config file see https://our.umbraco.org/projects/backoffice-extensions/azure-blob-storage-provider
One of the beauties of Umbraco is that it has so many hooks you can tap into and therefore bend it to your will, also great community support, looking at your requirements I would recommend you get an expert in to hand hold the build as it will save you alot of time also they will bring experience and show you how to bend it to your will :-}
Umbraco 7 questions (many-to-many / lookup tables and more).
Hello people,
I am looking into what kind of things I can build with Umbraco compared to using straight ASP.NET. I have tried to understand the available documentation, but the lacking of quality and mismanagement of the written content only gives small hints about what is possible and what is not.
1) Many to many relationships (in a graph where many items can refer to the same).
I want to create a repository system where everything can refer to everything bi-directional. E.g. like in a movie database: an actor can be the star in multiple movies, a movie can belong to multiple categories (action/drama). In a software system, you could have: a software package deployed on a host.
In both cases, searching for a genre should suggest the movies and searching for a moviestar should show the movies that he or she is in. I have seen that for 1 to many relationships I could use the Relations API. However I did not find a clear answer to if it can be used to design Many-to-many solutions... or if this is recommended in Umbraco or possible. Umbraco has a tree-structure that does not look suitable if all nodes are referring to each other as in the graph data structure.
2) Look up tables - adding new values in one place for universal values.
What is a common way to create a list of values that can be used to "group" things? I know I can use a Document Type as base, and create other Document Types under that so they get the value from its base. E.g. A list of pre-defined countries or departments.
3) Add tables in the Umbraco database and maybe integerate it in Umbraco.
It is OK to use the database that Umbraco uses, and add my own tables which will be used in the solution? Will it have a bad impact when upgrading Umbraco?
4) Case with sharing data (need to reference the same document unless the IsNotShared is not on).
E.g. in an Issue / Solution system. Let's say I have created a single software product which is used by four teams. When a group finds an issue and solution, they can add it to the software products' Issue / Solution System. Now, if one of the groups make changes so it is a now a more customized version, how it is possible for that group to add Issue / Solution items to the same software product, so THEY can see it, but not the other groups who share the other common features? Each group would have access to the same I/S system, but some of the other I/S tickets would only be viewable for the group who creates it (unless they want to show it).
It's hard for me to imagine a tree-structure in Umbraco that can do this... and new groups could be created on the fly...
5) Can the content node contain enough items for maybe a full-blown ticket system?
I'm not thinking about the performance of the backoffice GUI when it needs to show nodes, more if the display of data to the user on the frontend will be affected?
6) Visual Studio and the media library - avoid copying the media folder.
I can see that Umbraco saves images and flat files directly in the Umbraco folder. If more and more people add flat files and images to that folder it will take a long time to import the folder into Visual Studio. Are there any best practices for moving the media folder to another location and have Umbraco reference it? It would surely save some time when importing into VS.
This was a long post, I hope a friendly soul out there will have the time to give me an answer :)
Thomas,
1. There is a relations package and using the multi tree node picker you can relate things and do it bi directional its obviously not going to be as good as a graph db so if you really have loads of relations use a graph db, you have the data in umbraco so its easy for editors to edit then on publish using event handlers you can push that data to your graph db of choice im really a big fan of neo4j I have done any umbraco with it but Pete Duncanson who is quite popular in the umbraco world has done some prototyping.
2.You can do it with your own doc types however you can also do it csv files and use nupickers package, this will read the csv and give you a drop down we used it on a recent project to tag items with data like shoe size, hair colour etc.
3.You can add your own tables, upgrade will not trash those custom tables, also with upgrades db upgrades are done with migration scripts via peta poco.
4.This looks a bit complex so I do not fully understand it but if they create the issue you can add it to the product and using events set the permissions so this group can see and others cannot.
5. You can show the data on front end in paged manner? You can also on front end retrieve the data using Examine / Lucene searches so its lightening quick
6. There are amazon s3 providers that you can use and swap out the out of the box media provider, we are doing this on a current project as we have over 50gb of media so it all goes up into the cloud as blob storage. This is done by modifying the FileSystemProdiers.config file see https://our.umbraco.org/projects/backoffice-extensions/azure-blob-storage-provider
One of the beauties of Umbraco is that it has so many hooks you can tap into and therefore bend it to your will, also great community support, looking at your requirements I would recommend you get an expert in to hand hold the build as it will save you alot of time also they will bring experience and show you how to bend it to your will :-}
Regards
Ismail
Thanks for taking the time to answer my questions Ismail. It was very kind of you. It gave me some ideas on how to proceed.
is working on a reply...